Headlines this week trumpeted a new study produced by Nielsen and commissioned by Facebook which indicates they are "capable of delivering site reach at levels comparable to major TV networks". This follows on (coincidentally, we're sure) anonymously-sourced reports from earlier in the week that Facebook is set to price video ad units at between $1-2.5mm per day. The news-flow has served to conflate a number of factors and implicitly over-states the opportunity on hand for Facebook. While our model has long-assumed that Facebook will generate incremental revenue from online video, it will most likely be more modest than optimists might expect, and what they generate will not likely come from television budgets. Instead, we expect re-allocations of existing Facebook budgets or shifts of spending that would otherwise have gone to other media owners with online video inventory.
While Facebook reaches more people during the day than certain TV networks, this is a selective way of looking at the industry. Highlighting Facebook's dominant reach vs. individual networks during daytime wrongly implies that time-of-day is a first cut against which budgets are set. Perhaps they should be, but workflows associated with media buying (not to mention the distinct ways in which different media may influence consumers) dictate that the first cut of planning for large marketers is by medium. Digital media is separate from TV in this regard for many reasons: workflows are different, and lean-forward vs. lean-back environments mean they are generally viewed as different environments (not better or worse, just different), the means of assessment (how a budget is justified and how impressions are measured at the present time) are generally different, too. Comparing Facebook by itself to any one network is similarly moot: that a single media owner on one lower-reaching medium has more reach than single media-owners on higher reaching media is mostly irrelevant.
If Facebook's...er, Nielsen's implications were correct, advertisers would be more likely to shift their spending from TV to other publishers with online video inventory than to Facebook. Take Google's (GOOG, Hold) YouTube, for a start (notwithstanding that YouTube won't support Nielsen's OCR standard). Or video ad networks. Or home page take-overs involving video on virtually any portal. There are many ways to reach large audiences instantly on the web with video. Facebook's primary appeal, by contrast, has been the degree to which you can target target audiences vs. almost any alternative: an advertiser can often find more of a target there rather than anywhere else, but that doesn't necessarily mean that all of a sudden they want to use Facebook as a broad reach vehicle.
The pricing of units is, in a word, significant. Most digital budgets, when discretely negotiated, amount to tens of thousands, and occasionally hundreds of thousands of dollars. While a single large advertiser's annual commitment to Facebook may amount into the millions for annual campaign spending, the absolute commitment which Facebook is asking for is well-beyond the scope of what advertisers will pay in a single one-off buy. This leads us to suspect that over time Facebook will iterate its offering to include video assets which are priced at much lower costs and included in general ad sales packages. This is not a negative, per se: new product iteration is generally a positive thing, as it helps sales justify higher pricing for media. But even then, if it were capturing share it would come from budgets intended for Google's YouTube or online video networks rather than television.
If there is a single most important take-away from the PR effort associated with this news, it is that it may serve to highlight how "reach" can be more important than impressions or ratings, especially in context of our focus on why we have argued that television dominates the media budgets for most of the world's largest brands. Reach matters more than ratings from a planning perspective, and planners are more influential than buyers in allocating budgets across media. A slide in Nielsen's presentation on this topic highlights the right way to think about reach and media: once you start allocating more than 10-15% of a budget online, the amount of incremental reach you can produce starts to fall off substantially. This is why time and money won't necessarily ever equate to each other and why - given that most large brands are by now allocating 10-15% of their budgets to digital media - growth on the online advertising properties large brands favor has begun to slow significantly. Facebook can keep growing of course, but it will primarily occur as they take share of online budgets from existing segments of marketers and as they create new ad products for new segments of marketers.
Putting aside what a media plan says is optimal and the political challenges in implementing the necessary changes within advertisers' marketing organizations to execute on those optimal plans, another consideration is the cost to execute different kinds of media buys. It might cost 1.5% of the TV budget to actually buy TV, but 10% of a digital budget to buy digital media using conventional workflows. Agency contracts typically aren't structured in a manner which facilitates this, as marketers and their procurement teams often focus on the absolute amounts they pay their agencies even if it leads to inefficiencies in the media mix. We can argue that programmatic buying of digital media drives that cost down towards 2% and will help make marketers and agencies more indifferent in their media planning, although driving programmatic buying throughout the industry will take some time: most media owners would rather avoid selling their best inventory in this manner, and most buyers aren't yet ready to buy most media in this way, either. This past week's merger announcement between Publicis and Omnicom is important in this regard, as such changes are more likely than not to emerge as the new company will more likely invest the capital and develop the negotiation constructs necessary to make it happen faster than would otherwise have been the case. Statements from executives at Interpublic's Mediabrands this week around their intention to make 50% of all media buying programmatic is at minimum a reflection of the intentions that all agencies have towards these ends.
For a good illustration of how issues like the above play out elsewhere, consider: if marketers could identify the 20% of local markets that drive 80% of variations in outcomes, wouldn't they spend more money on local media? It's a potentially massively more efficient way to spend money, and one which marketers have had the capacity to execute against for many many years...and yet if anything marketers with nationally-skewed corporate structures have been shifting most of their spending towards national media over the same time. Consider that over the past ten years, national mass media owners' advertising revenues grew by 33% while local mass media owners' advertising revenues fell by 28% during the same time. Reasons unrelated to marketing efficiency usually have substantial sway in dictating changes in advertising spending choices across the economy.
Facebook will undoubtedly continue to grow the budgets it collects from the large advertisers who have video-based creative advertising units (and who represent perhaps a quarter to a third of the company's revenue base) by high single or low double digits much as it has over the past year or two. This is mostly a function of the mechanics of ad sales when a media owner is already one of the top suppliers of media to so many marketers. To justify these budget gains, Facebook needs to iterate and improve its offering for these advertisers on an ongoing basis, and a video ad product is one way to do this. It is usually the case that budgets shift generally in line with where audiences go within a medium, so to the extent that Facebook can capture increased share of digital media-based media time, that would lead to an incremental amount of revenue, too.
As before, the bulk of Facebook revenue growth comes, we think, from small businesses (which have doubled in number on Facebook, year-over-year) and e-commerce based advertisers (whose spending has doubled, according to the company). For a factor which causes us to meaningfully rethink our still very favorable forecasts for the growth trajectory of Facebook's ad revenue growth, we'd rather wait for a news release from Facebook on the trends impacting these cohorts instead.
Brian Wieser, CFA
Senior Research Analyst
Pivotal Research Group